Our goal is to understand further two aspects of gene expression: 1) the mechanisms that operate to 'open' or activate chromatin and 2) those that dictate how a particular gene within an active multi-gene cluster is chosen for expression. The working model evolving from studies of the human beta-globin cluster is that control elements upstream of the genes provide a general locus activation function (by an unknown mechanism). These upstream elements also increase the expression of the nearest available genes, with availability determined by the promoter. Human diseases with defects in each of these processes are known (eg. beta- thalassemia (Hispanic form) and hereditary persistence of fetal hemoglobin). Knowledge of these topics is part of the background needed for a rational approach to gene therapy. We previously demonstrated that the chicken betaA-globin gene and its 3' enhancer contain information sufficient to guarantee position-independent expression in transgenic mice. (This is quite different from the human beta-globin cluster where the beta-globin gene needed elements 35 kb to 50 kb upstream to guarantee transgene expression.) In the chick betaA-globin transgenics, the enhancer was essential for transgene expression. Promoter chromatin activation, as measured by DNase I hypersensitivity, correlated with transcription. We have now made mice containing the enhancer without the betaA-globin promoter, and are examining chromatin activation in these mice. We have previously identified pan-erythroid hypersensitive sites upstream of the beta-globin cluster. These sites were assayed for enhancer activity, with preliminary data demonstrating such activity for 2 of the sites. We are now characterizing these enhancers with respect to developmental stage specificity, cis elements, and trans factors. We are also examining their function in transgenic mice. A final project, in its early stages, is the production of transgenic mice carrying fragments of chicken genomic DNA containing all 4 chicken beta- globin genes, with and without the upstream hypersensitive sites and the betaA/epsilon enhancer. These mice will address the question of which control elements interact with which genes.